quant-metrics

Pure-math portfolio statistics for Rust. Decimal-backed, zero I/O, zero async, WASM-compatible.

Conventions match Bloomberg / MetaTrader / TradeStation where they exist (e.g. Sharpe uses linear risk-free approximation per Bloomberg, Sortino uses total-N denominator per Sortino & Price 1994, win rate excludes zero-PnL from wins).

Install

[dependencies]
quant-metrics = "0.4"

Statistics

Category	Functions
Returns	`total_return`, `annualized_return`, `cagr`
Risk-adjusted	`sharpe_ratio`, `sortino_ratio`, `calmar_ratio`, `information_ratio`
Drawdown	`max_drawdown`, `drawdown_series`, `max_drawdown_duration`, `recovery_time`
Risk	`value_at_risk` (VaR), `conditional_var` (CVaR)
Trade stats	`win_rate`, `profit_factor`, `avg_win`, `avg_loss`, `expectancy`
Composition	`compose`, `compose_mixed`, allocation methods (equal-weight, inverse-vol, HRP)
Cointegration	`pearson_correlation`, OLS hedge ratio, ADF test, Ornstein-Uhlenbeck half-life
Rolling	`RollingWindow` for returns, vol, sharpe, max-drawdown

Example

use quant_metrics::{sharpe_ratio, max_drawdown};
use rust_decimal_macros::dec;

let returns = vec![dec!(0.01), dec!(-0.02), dec!(0.03), dec!(0.005)];
let sharpe = sharpe_ratio(&returns, dec!(0.02), 252);  // rf=2%, daily
let dd = max_drawdown(&returns);

Audit

See AUDIT-VERDICT.md at the workspace root for formula correctness, numerical stability, and convention notes (reviewed by Claude Opus agent personas — not a human-conducted audit).

License